Clinical application of machine learning models in patients with prostate cancer before prostatectomy

Cancer Imaging

Table 2 Performance metrics for the five predictive models in the discovery and validation data sets. Error limits are +/- 1 standard deviation across 1000 CV splits

	Metric	Measurable ECE	Clinical	Clinical + Semantic	Clinical + Radiomic	Clinical + Semantic + Radiomic
Discovery	AUC	0.759	0.802 ± 0.149	0.887 ± 0.102	0.845 ± 0.121	0.880 ± 0.101
	Accuracy	0.871	0.790 ± 0.102	0.835 ± 0.092	0.811 ± 0.099	0.833 ± 0.092
	F1	0.679	0.560 ± 0.232	0.625 ± 0.225	0.602 ± 0.222	0.623 ± 0.228
Validation	AUC					0.928
	Accuracy					0.782
	F1					0.6
	p-value					2.3 × 10^− 7

ISSN: 1470-7330